- block: reorganize claim/release implementation
- block: make blkdev_get/put() handle exclusive access
- block: simplify holder symlink handling
- btrfs: close_bdev_exclusive() should use the same @flags as the matching open_bdev_exclusive()
- mtd: fix bdev exclusive open bugs in block2mtd::add_device()
- block: clean up bdev claim/release handling
- dm: lock bd_mutex when setting device size
- dm: use revalidate_disk to update device size after set_capacity
- multipath-tools libcheckers/directio.c libchec ...
- Antw: Q: md-RAID1 creation with I/O error (attempt to access beyond end of device)
- dm-kcopyd: Delayed unplug of the queues
- Remove BIO_RW_SYNCIO flag from kcopyd
- bad performance and system stalls when using dm-crypt
- CFQ and dm-crypt
- dm: use little-endian bit operations
- /usr/src/linux/drivers/md/bitmap.c: strange rounding lines 855-875
- Q: md-RAID1 creation with I/O error (attempt to access beyond end of device)
- failure in path between fc switch and storage: info request
- dm: use revalidate_disk to update device size after set_capacity
- dm-snap-mv update
- device mapper multipath and /sys/block/XXX/device/queue_dept
- unmapping luns doesn't work (good)
- Fix missing dm_put()
- scsi_dh_rdac: Add two new SUN devices to rdac_dev_list
- Make it possible to set a uuid if one was not set during DM_DEV_CREATE.
- block: eliminate potential for infinite loop in blkdev_issue_discard
- dm: use little endian bit operations
- ioctl HDIO_GETGEO to device mapper devpath
- device-mapper: Allow setting of UUID via rename if not already set
- workqueue: fix HIGHPRI handling in keep_working()
- Fix double free and use generic private pointer in per-cpu struct
- DM-CRYPT: Scale to multiple CPUs v3
- raid0 and discard/trim - current state?
- Q: Device mapper core and private biosets
- device-mapper: Allow setting of UUID via rename if not already set
- Make it possible to set a uuid if one was not set during DM_DEV_CREATE.
- Problem regarding splitting and merging of bio's in dm target
- Make it possible to set a uuid if one was not set during DM_DEV_CREATE.
- Linux 2.6.36-rc7
- Make it possible to set a uuid if one was not set during DM_DEV_CREATE.
- introduce dm-snap-mv
- New: 2.6.36-rc6 BUG at drivers/scsi/scsi_lib.c:1113
- dm log userspace add version number
- dm log userspace check additional error conditions
- dm-thinp: new device mapper target to thin provision storage
- dm-thinp: new device mapper target to thin provision storage
- block: eliminate potential for infinite loop in blkdev_issue_discard
- SMP aware kcryptd
- block: set the bounce_pfn to the actual DMA limit rather than to max memory
- scsi_dh_emc vs scsi_dh_alua in RHEL5.5
- block: set the bounce_pfn to the actual DMA limit rather than to max memory
- Is it deliberate that the device mapper is not passing the DISCARD ioctls
- idr_get_new_exact ?
- Is it deliberate that the device mapper is not passing the DISCARD ioctls
- dm: Fix possible NULL dereferences
- virtio-blk: put request that was created to retrieve the device id
- Help ! : device-mapper: reload ioctl failed: Invalid argument
- OT: device mapper development/testing.
- Combined storage tree
- multipath-tools/libmultipath dict.c
- dm: use cpu_relax when busy waiting for condition
- virtio-blk: put request that was created to retrieve the device id
- dm: convey that all flushes are processed as empty
- i_size misuses
- dm: convey that all flushes are processed as empty
- dm: convey that all flushes are processed as empty
- dm_crypt: very low performance of random IO operations
- jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier
- multipath-tools/libmultipath config.h dict.c s ...
- block, fs: replace HARDBARRIER with FLUSH/FUA, take#2
- jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier
- dm: fix locking context in queue_io()
- block: make __blk_rq_prep_clone() copy most command flags
- block: update documentation for REQ_FLUSH / REQ_FUA
- block: make sure FSEQ_DATA request has the same rq_disk as the original
- block: kick queue after sequencing REQ_FLUSH/FUA
- block: initialize flush request with WRITE_FLUSH instead of REQ_FLUSH
- block, fs: replace HARDBARRIER with FLUSH/FUA, take#2
- dm_crypt: very low performance of random IO operations
- doubt in installation of dm-ioband
- option to multipath to not modify the bindinfs file
- Discard support for dm-snap
- dm: implement REQ_FLUSH/FUA support for bio-based dm
- block: initialize flush request with WRITE_FLUSH instead of REQ_FLUSH
- dm: implement REQ_FLUSH/FUA support for request-based dm
- block: remove the WRITE_BARRIER flag
- dm: implement REQ_FLUSH/FUA support for request-based dm
- dm: relax ordering of bio-based flush implementation
- dm: implement REQ_FLUSH/FUA support for bio-based dm
- block: make __blk_rq_prep_clone() copy most command flags
- block, dm: finish REQ_FLUSH/FUA conversion, take#2
- block: QUEUE_FLAG_CLUSTER and no_cluster are NOT in sync
- block: Set stacked device's bounce_gfp same as that of the underlying device
- block: blk_phys_contig_segment() doesn't account for bounce pages
- imported patch propagate-bounce-gfp-to-stacked-drivers.patch
- imported patch sync-no_cluster-with-cluster-flag.patch
- multipath-tools/libmultipath dict.c parser.c p ...
- oblem with lvm and multipath on fedora 13
- option to multipath to not modify the bindinfs file
- block: update documentation for REQ_FLUSH / REQ_FUA
- jbd2: Modify ASYNC_COMMIT code to not rely on queue draining on barrier
- training mpath to discern between SCSI errors
- block, fs: replace HARDBARRIER with FLUSH/FUA
- block: remove the BLKDEV_IFL_BARRIER flag
- block: replace barrier with sequenced flush)
- block: remove the BH_Eopnotsupp flag
- block: remove the BLKDEV_IFL_BARRIER flag
- swap: do not send discards as barriers
- fat: do not send discards as barriers
- ext4: do not send discards as barriers
- jbd2: replace barriers with explicit flush / FUA usage
- jbd: replace barriers with explicit flush / FUA usage
- nilfs2: replace barriers with explicit flush / FUA usage
- reiserfs: replace barriers with explicit flush / FUA usage
- gfs2: replace barriers with explicit flush / FUA usage
- btrfs: replace barriers with explicit flush / FUA usage
- xfs: replace barriers with explicit flush / FUA usage
- block: pass gfp_mask and flags to sb_issue_discard
- lguest: replace VIRTIO_F_BARRIER support with VIRTIO_F_FLUSH support
- virtio_blk: drop REQ_HARDBARRIER support
- block: filter flush bio's in __generic_make_request()
- ide: remove unnecessary blk_queue_flushing() test in do_ide_request()
- block, fs: replace HARDBARRIER with FLUSH/FUA
- md: implment REQ_FLUSH/FUA support
- multipath-tools/multipath mpath_wait multipath ...
- DM manual failback doesn't work on RHEL5
- block: simplify queue_next_fseq
- lguest: replace VIRTIO_F_BARRIER support with VIRTIO_F_FLUSH support
- virtio_blk: drop REQ_HARDBARRIER support
- snapshots of mirror problems
- option to multipath to not modify the bindinfs file
- pvmove hangs
- virtio_blk: implement REQ_FLUSH/FUA support
- dm: implement REQ_FLUSH/FUA support
- md: implment REQ_FLUSH/FUA support
- lguest: replace VIRTIO_F_BARRIER support with VIRTIO_F_FLUSH/FUA support
- virtio_blk: implement REQ_FLUSH/FUA support
- block/loop: implement REQ_FLUSH/FUA support
- block: convert to REQ_FLUSH/FUA
- loader: parseCmdLineIp* takes just the value as an argument now.
- linux-2.6.35+ causes LVM to fail with " device-mapper: version ioctl failed: Inappropriate ioctl for device"
- linux-2.6.35+ causes LVM to fail with " device-mapper: version ioctl failed: Inappropriate ioctl for device"
- block: deprecate barrier and replace blk_queue_ordered() with blk_queue_flush()
- linux-2.6.35+ causes LVM to fail with " device-mapper: version ioctl failed: Inappropriate ioctl for device"
- block: replace barrier with sequenced flush
- block: replace barrier with sequenced flush
- block: replace barrier with sequenced flush
- block: replace barrier with sequenced flush
- fs, block: propagate REQ_FLUSH/FUA interface to upper layers
- block: use REQ_FLUSH in blkdev_issue_flush()
- fs, block: propagate REQ_FLUSH/FUA interface to upper layers
- block: implement REQ_FLUSH/FUA based interface for FLUSH/FUA requests
- block: rename barrier/ordered to flush
- block: rename blk-barrier.c to blk-flush.c
- block: drop barrier ordering by queue draining
- block: misc cleanups in barrier code
- block: remove spurious uses of REQ_HARDBARRIER
- block: deprecate barrier and replace blk_queue_ordered() with blk_queue_flush()
- block: kill QUEUE_ORDERED_BY_TAG
- block/loop: queue ordered mode should be DRAIN_FLUSH
- block: replace barrier with sequenced flush
- Fix traceback when using duplicate name for added/edited repo (#623080)
- device-mapper patches for 2.6.36
- multipath-tools/multipath multipath.conf.5
- dm-throttle: new device mapper target to throttle reads and writes
- [PATCH} dm-throttle: new device mapper target to throttle reads and writes
- scsi_dh_rdac: retry IO for 06/3f/03 in rdac_check_sense fn
- md updates for 2.6.36
- relaxed barriers
- Questions about dm-raid isw (device naming and partition table)
- How does dm-raid isw compare to other linux software raid solutions (e.g. mdadm)?
- dm: Fixes memory leak in setup_indexes function
- Performance impact of CONFIG_SCHED_MC? direct-io test case
- Performance impact of CONFIG_DEBUG? direct-io test case
- bio, fs: update READA and SWRITE to match the corresponding BIO_RW_* bits
- bio, fs: update READA and SWRITE to match the corresponding BIO_RW_* bits
- Use part instead of device in PartitionWindow.populate() (#575749)
- PartitionWindow.populate() fix
- : Fix for memory leak in setup indexes
- How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
- How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
- How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
- dm: support REQ_FLUSH directly
- relaxed cache flushes
- multipath: add fast_io_fail and dev_loss_tmo config parameters
- coda: rename REQ_* to CODA_REQ_*
- bio, fs: update RWA_MASK, READA and SWRITE to match the corresponding BIO_RW_* bits
- Fix up copying of firmware.
- Correct paths of things started by loader/init that have moved.
- Step 3 of merging installer images: No longer create install.img.
- makeinstimage is no longer used.
- instbin is no longer used.
- A couple minor changes to mk-images.
- The --loaderbin parameter to makeinitrd is unused. Kill it.
- Step 2 of merging installer images: Move most everything out of makeinitrd.
- Step 1 of merging installer images: Don't copy files into a new root.
- No longer do the bin -> usr/bin copy song and dance.
- The --initrdsize parameter to makeinitrd is unused. Kill it.
- A stern warning of things to come
- bio, fs: separate out bio_types.h and define READ/WRITE constants in terms of BIO_RW_* flags
- bio, fs: separate out bio_types.h and define READ/WRITE constants in terms of BIO_RW_* flags
- bio, fs: update READA and SWRITE to match the corresponding BIO_RW_* bits
- scsi_dh_rdac: retry IO for 06/3f/03 in rdac_check_sense fn
- multipath: add fast_io_fail and dev_loss_tmo config parameters
- scsi_dh : Couple of fixes for scsi device handlers
- Add support for dm-mod module autoloading
- scsi_dh : increment the refcounts while calling activate
- scsi_dh_rdac : Adding MODULE VERSION for rdac device handler
- scsi_dh : release the refcounts in scsi_dh_rdac handler when activate is complete
- scsi_dh : release the refcounts in scsi_dh_hp_sw handler when activate is complete
- scsi_dh : release the refcounts in scsi_dh_emc handler when activate is complete
- scsi_dh : release the refcounts in scsi_dh_alua handler when activate is complete
- scsi_dh : increment the refcounts while calling activate
- scsi_dh : rename and export __detach_handler
- scsi_dh : Couple of fixes for scsi device handlers
- RAID/block regression starting from 2.6.32, bisected
- dm stripe: move sector translation to a function
- dm stripe: optimize sector division
- dm: use dm_target_offset in targets
- How to associate pci id with wwpn for an hba
- dm-multipath: typo, strncat instead of strncpy
- dm stripe: enable discard support
- dm-stripe: optimize sector division
- dm-stripe: discard support
- dm-stripe: move sector translation to a function
- BIO_RW_SYNCIO
- rq-based mpath failed path on SCSI target failure (malformed discard)
- rq-based mpath failed path on SCSI target failure (malformed discard)
- dm-dirty-log: allow log size to be different from target size.
- The DM part of dm-raid45
- dm-raid456: add support for setting IO hints.
- dm-raid456: switch to use dm_dirty_log for tracking dirty regions.
- dm stripe: enable efficient discard support
- block: update request stacking methods to support discards
- dm: use common __issue_target_request for flush and discard support
- dm: remove the DM_TARGET_SUPPORTS_DISCARDS feature flag
- dm: enable discard support for more targets
- block: update request stacking methods to support discards
- dm stripe: enable efficient discard support
- dm mpath: enable discard support
- block: update request stacking methods to support discards
- dm delay: enable discard support
- dm error: return error for discards too
- dm zero: silently drop discards too
- dm: split discard requests on target boundaries
- dm: factor max_io_len for code reuse
- dm: use common __issue_target_request for flush and discard support
- dm: remove the DM_TARGET_SUPPORTS_DISCARDS feature flag
- dm: introduce num_discard_requests in dm_target structure
- dm: rename map_info flush_request to target_request_nr